document resume ed 379 339 tm 022 723 author melancon ... · the ppdsq (thompson, 1994) consists of...
TRANSCRIPT
DOCUMENT RESUME
ED 379 339 TM 022 723
AUTHOR Melancon, Janet G.; Thompson, BruceTITLE An Adjectival Self-Description Checklist Evaluating
Myers-Briggs Type Indicator (MBTI) Types: Concurrentand Construct Score Validity.
PUB DATE Nov 94NOTE 36p.; Paper presented at the Annual Meeting of the
Mid-South Educational Research Association(Nashville, TN, November 1994).
PUB TYPE Reports Research/Technical (143)Speeches /Conference Papers (150)
EDRS PRICE MF01/PCO2 Plus Postage.DESCRIPTORS Check Lists; *College Students; Higher Education;
*Measurement Techniques; *Personality Measures;Personality Traits; Reliability; *Scores; *Test Use;Validity
IDENTIFIERS Adjective Scales; Confirmatory Factor Analysis;Exploratory Factor Analysis; LISREL Computer Program;Myers Briggs Type Indicator; Personal PreferencesSelf Description Quest
ABSTRACTThis study was conducted to evaluate whether the
adjectival self-description checklist may provide a viable method ofquickly obtaining initial personality type information. The PersonalPreferences Self-Description Questionnaire (PPDSQ) was administeredto more than 420 college students, and data were analyzed usingclassical reliability analysis and both exploratory and LISRELconfirmatory factor analyses. Students also took the Myers BriggsType Indicator (MBTI). Results generally supported a conclusion thatPPSDQ scores are reasonably reliable and valid. Thirteen tables andtwo figures present details of the statistical analyses. An appendixpresents some descriptive MBTI statistics. (Contains 21 references.)(Author/SLD)
***********************************************************************
Reproductions supplied by EDRS are the best that can be madefrom the original document.
***********************************************************************
enCR ppsdg2ms.wp1 11/4/94McrN
M U.S. DEPARTMENT OF EDUCATION°Noce of Educational Research and Improvement
EOU TIONAL RESOURCES INFORMATIONCENTER (ERIC)
This document has been reproduced asreceived from the person Or organizationoriginating it
0 Minor changes have been made to imptOvereproduction Quality
PointS of view Or opinions staled in this docu-ment do not necessarily represent officialMMposmonormhCy
'PERMISSION TO REPRODUCE THISMATERIAL HAS BEEN GRANTED BY
hvomio.565/0
TO THE EDUCATIONAL RESOURCESINFORMATION CENTER (ERIC).-
AN ADJECTIVAL SELF-DESCRIPTION CHECKLISTEVALUATING MYERS-BRIGGS TYPE INDICATOR (MBTI) TYPES:
CONCURRENT AND CONSTRUCT SCORE VALIDITY
Janet G. Melancon Bruce Thompson
Loyola University Texas A&M University 77843-4225and
Baylor College of Medicine
BEST COPY AVAILABLE
Paper presented at the annual meeting of the Mid-SouthEducational Research Association, Nashville, TN, November 10, 1994.
2
Abstract
The study was conducted to evaluate whether the adjectival self-
description checklist may provide a viable method of quickly
obtaining initial personality type information. The Personal
Preferences Self-Description Questionnaire (PPDSQ) was administered
to more than 420 college students, and data were analyzed using
classical reliability analysis and both exploratory and LISREL
confirmatory factor analyses. Results generally supported a
conclusion that PPSDQ scores are reasonably reliable and valid.
3
The Myers-Briggs Type Indicator (MBTI) is one of the most
frequently used measures of personality, as noted in various
reviews (cf. Thompson & Ackerman, 1994). This has been the case
for at least two reasons. First, unlike many personality measures,
the MBTI focuses on normal variations in personality, and because
more people have normal as against abnormal personality, the
measure may be useful with more people than measures of
psychopathology would be. Second, many counselors find that the
MBTI has enormous "face validity" for clients, i.e., that clients
understand the concepts implicit in the measure, tend to agree with
important aspects of type characterizations, and find the
information to be useful, free of value judgments, and non-
threatening.
McCaulley (1990) provides a concise and informed overview of
the MBTI, its history, and its uses. The forms of the MBTI were
developed over at least four decades.. Initial work was done by
Katherine C. Briggs and her daughter, Isabel Briggs Myers. Mary H.
McCaulley also made numerous contributions, and worked closely with
Isabel in projects such as the writing of the comprehensive MBTI
manual (Myers & McCaulley, 1985), which was published subsequent to
Myers' death in May, 1980.
The MBTI was developed with some grounding in the basic
precepts of Carl G. Jung's theory of psychological functions and
types. The theory presumes that "...much of the seemingly random
variation in behavior is actually quite orderly and consistent,
being due to basic differences in the way individuals prefer to use
1
4
their perception and judgment" (Myers & McCaulley, 1985, p. 1).
The MBTI is designed to measure four dimensions: Extraversion-
Introversion, Sensation-iNtuition, Thinking-Feeling, and Judgment-
Perception. In conventional usage, continuous scores are computed
on each dimension for each preference pole of the dimension (e.g.,
both Extraversion and Introversion on the EI dimension), and
persons are "typed" based on which style within each dimension is
preferred. Each individual is then classified into one of the 16
types formed from all possible combinations of the four scales,
e.g., ENTJ, ISTP,. and ENFP.
Unlike the other three dimensions, the JP construct is
implicit (rather than explicit) within Jung's theory.
Theoretically, people do have a general rank-order preference for
the four mental processes or functions of Sensing, iNtuition,
Thinking, and Feeling. Myers reasoned that JP scores--when taken
together with EI scores--would point to a person's dominant,
auxiliary, tertiary, and inferior functions (see McCaulley, 1990;
Myers & McCaulley, 1985).
Persons with a preference for Judging most show the world in
their public persona or public face either Thinking or Feeling,
depending upon their preferences on the TF scale. Persons with a
preference for Perceiving have either Sensing or iNtuition as the
main function in their public persona. Persons with a preference
for Extraversion show the world their dominant function as part of
their public persona, while persons with a preference for
Introversion show the world their second-most prefered function,
2
i.e., their auxiliary. Thus, INTJ's show the, world most often a
preference for Thinking, but actually their most-prefered
function--their dominant--is iNtuition. ENTJ's, on the other hand,
most show the world Thinking, and Thinking is actually the dominant
for the extraverted ENTJ's.
MBTI items are forced-choice in nature and consist of paired
statements, one from either preference pole on one of the four
scales. The MBTI was designed for use with older adolescents and
adults in the normal population. Most forms of the measure have
roughly 100 scored items. Previous factor analytic investigations
of MBTI data have generally been supportive of a conclusion that
the instrument yields scores measuring the intended constructs
(e.g., Thompson & Borrello, 1986).
A pair of studies reported in the Manual (Myers & McCaulley,
1985) by Carskadon used self-estimate of type as a validity
measure. When subjects were asked to choose the type description
that best suited them, their actual MBTI-tested type was chosen to
a statistically significant degree more often than other types.
These studies prompted us to explore the utility of a short-
form measure of type preferences developed by Thompson (1994); this
measure is an adjectival self-description checklist--the Personal
Preferences Self-Description Questionnaire (PPDSQ). The MBTI
itself includes several items involving adjectival self-description
and this, taken with the previous research involving self-
estimation of type, together suggest that adjectival self-
description may provide a sufficient basis with which to
3
6
tentatively identify type.
The purpose of our present study was to explore the
reliability of PPSDQ scores, the concurrent validity of PPSDQ
scores in relation to MBTI continuous scores, and the construct
validity of PPSDQ scores. The study was conducted as a second step
in an iterative sequence of PPSDQ test revisions and refinements,
building on the previous work reported by Thompson and Stone
(1994).
Method
Subiects
We administered MBTI Form G and the PPDSQ self-descriptive
adjectival checklist to 420 college students enrolled in a private
university located in the southern United States in an urban
setting. There were more females (n1=273; 65.0%) than males
(n2=147; 35.0%) in our sample. The mean age of the sample was 23.82
(SD=9.58).
Instrumentation
The PPDSQ (Thompson, 1994) consists of 59 scored adjective-
pairs posited to mark each of the four dimensions of personality
measured by the MBTI. Roughly half the PPSDQ items measuring each
of the four constructs were reversed so as to minimize response
set. For example, item 1 ("Quiet-Expressive") measures EI, but the
Introversion adjective ("Quiet") is presented first within the
pair. Item 6 ("Social-Private") also measures EI, but the
Extraversion adjective ("Social") is presented first within this
adjective pair.
4
Each adjective pair is presented as a semantic differential
scale. A Likert scale ("1" to "7") is presented between each pair
of adjectives, and subjects circle the number that represents which
adjective best describes them. Thus, unlike the MBTI which uses an
"ipsative" or forced-choice response format, the PPDSQ uses a
"normative" or non-forced-choice response format.
Results
Presumptions Underlying Analytic Choices
In the present study the primary analyses involved classical
reliability statistics and principal components analyses. Prior to
elaborating these results, some discussion of the presumptions
underlying our major analytic methods seems warranted.
The Nature of Reliability. Unlike many researchers, we
consciously recognized that reliability is a characteristic of
scores or data in hand, and generally ought to be investigated for
every given data set. Many authors present this view, but paradigm
influences constrain some researchers from integrating this
presumption into their actual analytic practice (Thompson, in
press).
For example, Rowley (1976, p. 53, emphasis added) noted that,
"It needs to be established that an instrument itself is neither
reliable nor unreliable.... A single instrument can produce scores
which are reliable, and other scores which are unreliable." And
Crocker and Algina (1986, p. 144, emphasis added) argued that,
"...A test is not 'reliable' or 'unreliable.' Rather, reliability
is a property of the scores on a test for a particular group of
5
examinees."
In another widely respected text, Gronlund and Linn (1990,
78, emphasis in original) correctly noted,
Reliability refers to the results obtained with an
evaluation instrument and not to the instrument
itself.... Thus, it is more appropriate to speak of
the reliability of the "test scores" or of the
"measurement" than of the "test" or the
"instrument."
And Eason (1991, p. 84, emphasis added) argued that:
Though some practitioners of the classical
measurement paradigm (incorrectly] speak of
reliability as a characteristic of tests, in fact
reliability is a characteristic of data, albeit data
generated on a given measure administered with a
given protocol to given subjects on given occasions.
The sample itself impacts the reliability of scores.
Reliability is driven by variance--typically greater scores
variance leads to greater score reliability, and so more
heterogeneous samples often lead to more variable scores, and thus
to higher reliability. If the test could be reliable, score
reliability would logically not be influenced by to whom the test
was administered. Obviously, this is not the case!
The same measure, when administered to more heterogenous or
more homogeneous sets of subjects, will yield scores with differing
reliability. As Dawes (1987, p. 486) observed, "Because
P
6
reliability is a function of sample as well as of instrument, it
should be evaluated on a sample from the intended target
population--an obvious but sometimes overlooked point."
Our shorthand ways of speaking (e.g., language saying "the
test is reliable") can itself'cause confusion and lead to bad
practice. As Pedhazur and Schmelkin (1991, p. 82, emphasis in
original) observed, "Statements about the reliability of a measure
are... inappropriate and potentially misleading." These
telegraphic ways of speaking can be problematic, if we come
unconsciously to ascribe truth to our literal shorthand, rather
than recognize that our jargon is sometimes telegraphic and not
literally true. As Thompson (1992, p. 436) emphasizes:
This is not just an issue of sloppy speaking--the
problem is that sometimes we unconsciously come to
think what we say or what we hear, so that sloppy
speaking does sometimes lead to a more pernicious
outcome, sloppy thinking and sloppy practice.
The Utility of Principal Components Analyses for Informing
Judgments Regarding Construct Validity. With respect to using
factor analysis to help judge score validity, many researchers
acknowledge the prominent role that factor analysis can play in
efforts to establish construct validity. For example, Nunnally
(1978, p. 111) noted that, historically, "construct \lidity has
been spoken of ar [both] 'trait validity' and 'factorial
validity.'"
Similarly, Gorsuch (1983, p. 350) noted that, "A prime use of
7
10
factor analysis has been in the development of both the operational
constructs for an area and the operational representatives for the
theoretical'constructs." In short, "factor analysis is intimately
involved with questions of validity.... Factor analysis is at the
heart of the measurement of psychological constructs" (Nunnally,
1978, pp. 112-113).
But analysts differ quite heatedly over the utility of
principal components as compared to common or principal factor
analysis. For example, an entire special issue on this controversy
was recently published in the 1992 volume of Multivariate
Behavioral Research. The difference between the two approaches
involves the entries used on the diagonal of the correlation matrix
that is analyzed--principal components analysis uses ones on the
diagonal while common factor analysis uses estimates of
reliability, usually estimated through an iterative process.
The two methods yield increasingly more equivalent results as
either (a) the factored variables are more reliable or (b) the
number of variables being factored is increased. Snook and Gorsuch
(1989, p. 149) explained this second point, noting that "As the
number of variables decreases, the ratio of diagonal to off-
diagonal elements also decreases, and therefore the value of the
communality has an increasing effect on the analysis." For
example, with 10 variables the 10 diagonal entries in the
correlation matrix represent 10% (10 / 100) of the 100 entries in
the matrix, but with 100 variables the diagonal entries represent
only 1% (100 / 10,000) of the 10,000 matrix entries. Gorsuch
8
11
(1983) suggested that with 30 or more variables the differences
between solutions from the two methods are likely to be small and
lead to similar interpretations.
Anylsis #1: Score Reliability (v=59+8 and v=59)
We first computed classical, corrected, item discrimination
(i.e., is between scores on each item -- potentially ranging from
"1" to "7"--and scores on all the remaining items in each of the
four scales--potentially ranging from "v" [17 items x 1] to "v times
7") and scale alpha coefficients (cf. Thompson & Levitov, 1985).
For each scale we conducted analyses both (a) for PPSDQ scale items
and the 2 relevant continuous scores from the MBTI (e.g., the
Extraversion and the Introversion scores from the EI scale) and (b)
for only PPSDQ items. The first analysis uses MBTI scores to help
delineate or "mark" the construct measured in a given analysis.
These analyses are reported in Tables 1 through 8.
INSERT TABLES 1 THROUGH 8 ABOUT HERE.
Anylsis #2: Factor Structure (v=59+8 and v=59)
Next, we extracted four principal components from the inter-
item correlation matrix and rotated these factors to the varimax
criterion. In this initial factor analysis we used both scores on
the 59 PPSDQ items plus continuous scores on each pair of scores
for each of the four scales from the MBTI (i.e., MBTI scores on E,
I, S, N, T, F, J, and P). Thus, this analysis involved 67 (59+8)
items. Figure 1 presents the "scree" plot of the eigenvalues
associated with the factors prior to rotation (Thompson, 1989).
9
Table 9 presents the factor pattern/structure matrix rotated to the
varZmax criterion. The items in the table are sorted into the four
scales (i.e., SN, TF, EI, and JP) presumed to be measured by the
PPSDQ.
INSERT FIGURE 1 AND TABLE 9 ABOUT HERE.
Figure 2 presents the "scree" plot for the related analysis
involving only the 59 PPSDQ items. Table 10 presents the factor
pattern/structure matrix rotated to the varimax criterion. Factor
scores were also computed for each subject on each factor, for use
in concurrent validity analyses.
INSERT FIGURE 2 AND TABLE 10 ABOUT HERE.
Analysis #3: Confirmatory Factor Analysis of Scale Scores (v=12)
Confirmatory analyses were also conducted using covariance
structure analyses. The bivariate correlation matrix involving (a)
the 8 MBTI scale scores and (b) the 4 summated scale scores
computed by adding item responses on the items defining each PPDSQ
scale was the basis for these LISREL analyses (JOreskog & Sorbom,
1989).
The a priori model positing the existence of four correlated
factors yielded a x2 of 424.00 (df = 48; noncentrality parameter =
424.00 48 = 376.00; 376.00/48 = 7.83). The LISREL goodness-of-
fit index (GFI) was .76. The comprative baseline for the analysis
was a null model positing 12 uncorrelated factors and no
measurement error; this model yielded a x2 of 5,153.24 (df = 66;
10
13
noncentrality parameter = 5,153.24 66 = 5,087.24; 5,087.24/66 =
77.08). The LISREL goodness-of-fit index (GFI) for the null model
was .24. The test of the theoretical model is reported in Table
11.
INSERT TABLE 11 ABOUT HERE.
Analysis #4: Concurrent Validity of Scale and Factor Scores
The last analysis involved computed concurrent validity
coefficients on the 8 MBTI continuous scale scores, the 4 PPSDQ
factor scores computed as part on Analysis #2, and the 4 PPSDQ
summated scores computed by adding a subject's responses to all the
items on a given scale. These analyses are reported in both Tables
12 and 13.
INSERT TABLES 12 AND 13 ABOUT HERE.
Discussion
As is always the case, no one study taken alone should be
overinterpreted. As Neale and Liebert (1986, p. 290) observed:
No one study, however shrewdly designed and
carefully executed, can provide convincing support
for a causal hypothesis or theoretical statement...
Too many possible (if not plausible) confounds,
limitations on generality, and alternative
interpretations can be offered for any one
observation. Moreover, each of the basic methods of
research (experimental, correlational, and case
11
14
study) and techniques of comparison (within- or
between-subjects) has intrinsic limitations. How,
then, does social science theory advance through
research? The answer is, by collecting a diverse
body of evidence about any major theoretical
proposition.
Nevertheless, some tentative conclusions can be offered, based
on these results. First, as indicated in Table 1 through 8, it is
possible to derive scores from the PPSDQ that have reasonable
internal consistency. The most problematic of the four scales,
from this perspective, is the JP scale.
Differential scale performance for JP is indicated by the
remaining analyses as well. For example, in the exploratory factor
analysis reported in Tables 9 and 10, the JP scale tended to merge
with the SN scale. In the confirmatory factor analysis reported in
Table 11, the JP and SN scales were most highly correlated (r =
+.515; r2 = 26.5%), while all the other pairs of factors had
interfactor correlation coefficients ranging at the extremes from
1.0861 to 1.2621 (i.e., r2 values ranging from 0.7% to 6.95%). And
in the Pearson r and Spearman rho matrices presented in Tables 12
and 13, although PPSDQ Judging-Perceiving scores were highly
correlated with MBTI J and P continuous scores (e.g., r = -.6951
and r +.6911, respectively), PPSDQ Judging-Perceiving scores were
also highly correlated with MBTI J and P continuous scores (e.g.,
r = -.4165 and r +.4250, respectively).
It may be that more adjective self-description semantic-
12
differential items are needed to measure the JP scale. However,
taken together with previous results (Thompson & Stone, 1994),
these results may alternatively suggest that the JP scale is harder
to measure using semantic differential items. Items consisting of
complete sentences may be needed to elaborate this construct. These
alternatives remain to be explored in future research.
In summary, results in the present study were generally
favorable regarding at least three of the four constructs presumed
to be measured by the PPSDQ. Additional items need to be
formulated to tap the fourth (JP) dimension. Further research
using the original PPSDQ items together with additional items would
allow both replication of present and previous results, as well as
exploration of improvements resulting from use of an additional set
of JP items.
13
16
References
Crocker, L., & Algina, J. (1986). Introduction to classical and
modern test theory. New York: Holt, Rinehart and Winston.
Dawes, R.V. (1987). Scale construction. Journal of Counseling
Psychology, 34, 481-489.
Eason, S. (1991). Why generalizability theory yields better results
than classical test theory: A primer with concrete examples.
In B. Thompson (Ed.), Advances in educational research:
Substantive findings, methodological developments (Vol. 1, pp.
83-98). Greenwich, CT: JAI Press.
Gorsuch, R.L. (1983). Factor analysis (2nd ed.). Hillsdale, NJ:
Erlbaum.
Gronlund, N.E., & Linn, R.L. (1990). Measurement and evaluation in
teaching (6th ed.). New York: Macmillan.
Joreskog, K.G., & Sorbom, D. (1989). LISREL 7: A guide to the
program and applications (2nd ed.). Chicago: SPSS.
McCaulley, M.H. (1990). The Myers-Briggs Type Indicator: A measure
for individuals and groups. Measurement and Evaluation in
Counseling and Development, 22, 181-195.
Myers, I.B. & McCaulley, M.H. (1985). Manual: A guide to the
development and use of the Myers-Briggs Type Indicator. Palo
Alto, CA: Consulting Psychologists Press.
Neale, J.M., & Liebert, R.M. (1986). Science and behavior: An
introduction to methods of research (3rd ed.). Englewood
Cliffs, NJ: Prentice-Hall.
Nunnally, J.C. (1978). Psychometric theory (2nd ed.). New York:
14
17
McGraw-Hill.
Pedhazur, E.J., & Schmelkin, L.P. (1991). Measurement, design, and
analysis: An integrated approach. Hillsdale, NJ: Erlbaum.
Rowley, G.L. (1976). The reliability of observational measures.
American Educational Research Journal, 13, 51-59.
Snook, S.C., & Gorsuch, R.L. (1989). Component analysis versus
common factor analysis: A Monte Carlo study. Psychological
Bulletin, 106, 148-154.
Thompson, B. (1989). Prerotation and postrotation eigenvalues
shouldn't be confused: A reminder. Measurement and Evaluation
in Counseling and Development,.22(3), 114-116.
Thompson, B. (1992). Two and one-half decades of leadership in
measurement and evaluation. Journal of Counseling and
Development, 70, 434-438.
Thompson, B. (1994). Personal Preferences Self-Description
Questionnaire (PPSDO). College Station, TX: Author.
Thompson, B. (in press). Guidelines for authors. Educational and
Psychological Measurement.
Thompson, B., & Ackerman, C. (1994). Review of the Myers-Briggs
Type Indicator. In J. Kapes, M. Mastie & E. Whitfield (Eds.),
A counselor's guide to career assessment instruments (3rd ed.,
pp. 283-287). Alexandria, VA: American Counseling Association.
Thompson, B., & Borrello, G.M. (1986). Second-order factor
structure of the MBTI: A construct validity assessment.
Measurement and Evaluation in Counseling and Development, 18,
148-153.
15
is
Thompson, B., & Levitov, J.E. (1985). Using microcomputers to score
and evaluate test items. Collegiate Microcomputer, 3, 163-168.
Thompson, B., & Stone, E. (1994, January). Concurrent validity of
scores from an adjectival self-description checklist in
relation to Myers-Briggs Type Indicator (MBTI) scores. Paper
presented at the annual meeting of the Southwest Educational
Research Association, San Antonio, TX. (ERIC Document
Reproduction Service No. ED forthcoming)
Table 1Classical Reliability Statistics for
Extraversion-Introversion Scale (n=420, v=17)
SCALEMEANIF ITEMDELETED
SCALEVARIANCEIF ITEMDELETED
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
MIXERLON 46.1747 602.3110 .7376 .8388SOCPRIVA 45.7057 598.6838 .7113 .8383XINTREXT 45.7700 602.3286 .6953 .8393XSILENGA 45.4083 608.2712 .6415 .8412PERSNSHY 46.2319 601.0915 .6718 .8393XQUIETEX 46.2366 608.0975 .6315 .8413GREGARTI 45.8092 622.8944 .5176 .8455CONGRECL 46.1057 623.6358 .5205 .8456FRIEDIST 47.0975 625.2314 .4981 .8462XSOLIAMI 46.0742 624.5341 .5146 .8459EXUBSERE 45.2342 625.1357 .4413 .8469XSTILLAN 46.2700 628.3960 .4230 .8476XREFLECA 45.7019 627.8653 .3541 .8488APPROACH 45.9057 631.3995 .2974 .8503XTERSEWO 45.5754 645.2465 .1773 .8534XEXTRAVE 63.4676 381.2318 .8843 .8276INTROVER 36.1223 372.0286 .8573 .8379a = 0.8516
Note. MBTI continuous variables are italicized.
Table 2Classical Reliability Statistics for
Extraversion-Introversion Scale (n=420, v=15)
SCALEMEANIF ITEMDELETED
SCALEVARIANCEIF ITEMDELETED
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
MIXERLON 47.1533 154.6882 .7298 .8547SOCPRIVA 46.6842 153.3749 .6867 .8562XINTREXT 46.7485 154.6072 .6871 .8564XSILENGA 46.3869 157.8355 .6263 .8597PERSNSHY 47.2104 152.4019 .7025 .8552XQUIETEX 47.2152 157.3937 .6251 .8597GREGARTI 46.7878 165.1652 .5064 .8656CONGRECL 47.0842 164.1198 .5543 .8637
FRIEDIST 48.0761 165.3362 .5192 .8652XSOLIAMI 47.0527 164.6010 .5485 .8640EXUBSERE 46.2128 165.7207 .4427 .8685XSTILLAN 47.2485 167.2928 .4297 .8689XREFLECA 46.6804 168.2260 .3215 .8751APPROACH 46.8842 168.0372 .3104 .8761XTERSEWO 46.5539 176.3760 .1706 .8798a = 0.8727
Table 3Classical Reliability Statistics forSensing-Intuition Scale
SCALE SCALEMEAN VARIANCEIF ITEM IF ITEMDELETED DELETED
(n=420, v=16)
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
TRADCREA 56.3901 487.2768 .5016 .7882PRECIMAG 56.5491 484.4952 .5474 .7865XINVENOR 56.9277 485.3414 .5483 .7867PLANVISI 56.5563 490.3602 .5197 .7888CONCLEXP 56.1444 497.7096 .4446 .7923XINSIGHT 56.4848 490.2759 .4843 .7894XDIVERCO 56.5405 490.5354 .4695 .7898REALINTU 57.5214 494.9152 .3872 .7927XDIVERPR 56.7410 489.9822 .4851 .7893XCONCEPR 57.7958 500.9453 .3385 .7951DIRECTIN 56.7720 500.9396 .3881 .7941PRACTHEO 57.8269 503.1276 .3143 .7961XVARIREP 55.8944 502.4458 .3549 .7950XINQUICR 56.4705 504.3441 .2959 .7967XSENSING 73.8086 259.4891 .8153 .7811INTUITIO 49.2063 314.6216 .7908 .7536a = 0.8001
Note. MBTI continuous variables are italicized.
Table 4Classical Reliability Statistics forSensing-Intuition Scale
SCALE SCALEMEAN VARIANCEIF ITEM IF ITEMDELETED DELETED
(n=420, v=14)
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
TRADCREA 57.0544 115.6057 .4639 .8070PRECIMAG 57.2134 112.3761 .5704 .7986XINVENOR 57.5920 112.2504 .5901 .7973PLANVISI 57.2205 115.1862 .5500 .8010CONCLEXP 56.8086 119.9100 .4377 .8091XINSIGHT 57.1491 114.7054 .5225 .8026XDIVERCO 57.2048 113.8563 .5362 .8014REALINTU 58.1857 119.3624 .3454 .8163XDIVERPR 57.4052 112.7259 .5822 .7980XCONCEPR 58.4601 123.7873 .2527 .8219DIRECTIN 57.4363 120.8139 .4039 .8113PRACTHEO 58.4911 124.9905 .2241 .8235XVARIREP 56.5586 120.5781 .4027 .8113XINQUICR 57.1348 120.6542 .3624 .8142a = 0.8196
1821
Table 5Classical Reliability Statistics forThinking-Feeling Scale
SCALE SCALEMEAN VARIANCE
IF ITEM IF ITEMDELETED DELETED
(n=420, v=23)
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETEDFACTCOMP 90.8783 647.3906 .6542 .8262XTENDERR 91.3079 651.8917 .5602 .8281XFEELTHI 91.3351 648.7238 .5286 .8280XKINDANA 90.8899 651.8594 .5366 .8284STRICTFO 90.5608 659.8229 .4988 .8301DISPASEM 90.3090 660.3387 .4928 .8303SKEPTRUS 91.2304 653.7754 .4711 .8296XEMPATHL 92.1162 659.0428 .4306 .8310LOGICHUM 91.4018 658.5109 .4477 .8306XLIGHTHE 90.8470 662.4211 .4403 .8313XGULLSUS 92.3113 667.0988 .3877 .8326XCARICOO 90.3985 665.2417 .3931 .8323XACCEPDI 90.6590 662.8569 .4146 .8317
XRECEPTS 91.5375 666.1008 .3584 .8330EVALNONJ 91.6899 665.5734 .3591 .8329XSYMPATH 92.0351 671.7724 .2668 .8353JUSTHARM 91.7113 669.8833 .2852 .8348EVALOPEN 91.0804 670.5456 .2949 .8346PRINCIPL 91.2182 668.4790 .2945 .8345IMPERPER 90.2280 676.7697 .2947 .8350XSENSUAL 90.9904 678.9460 .2036 .8369XTHINKIN 106.5970 386.8658 .8239 .8337FEELING 85.5209 487.3821 .7781 .8094a = 0.8373
Note. MBTI continuous variables are italicized.
19
Table 6Classical Reliability Statistics forThinking-Feeling Scale
SCALE SCALEMEAN VARIANCEIF ITEM IF ITEMDELETED DELETED
(n=420, v=21)
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
FACTCOMP 91.4640 218.5737 .6521 .8178XTENDERR 91.8936 222.1404 .5326 .8226XFEELTHI 91.9209 221.1924 .4811 .8245XKINDANA 91.4756 220.8364 .5372 .8221STRICTFO 91.1465 225.5817 .4999 .8244DISPASEM 90.8947 226.3701 .4817 .8252SKEPTRUS 91.8161 221.7894 .4713 .8250XEMPATHL 92.7019 227.1178 .3835 .8292LOGICHUM 91.9875 225.6744 .4252 .8272XLIGHTHE 91.4328 226.4460 .4546 .8261XGULLSUS 92.8970 231.0177 .3584 .8301XCARICOO 90.9843 228.7260 .3911 .8288XACCEPDI 91.2447 226.2817 .4361 .8268XRECEPTS 92.1232 229.1622 .3557 .8304EVALNONJ 92.2756 228.1742 .3705 .8297XSYMPATH 92.6209 235.8533 .1935 .8381JUSTHARM 92.2970 231.1845 .2830 .8341EVALOPEN 91.6661 230.5495 .3163 .8322PRINCIPL 91.8040 228.5244 .3279 .8320IMPERPER 90.8137 233.9129 .3391 .8309XSENSUAL 91.5761 235.3526 .2292 .8358a = 0.8352
202
Table 7Classical Reliability Statistics forJudging-Perceiving Scale
SCALE SCALEMEAN VARIANCEIF ITEM IF ITEMDELETED DELETED
(n=420, v=11)
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
XFLEXORG 35.6095 452.6252 .5669 .7689PROMPTFR 35.3557 449.2314 .5830 .7671XRANDSEQ 35.7795 456.0741 .5535 .7708TIMELYRE 35.3581 454.8683 .5342 .7706XIMPETTA 36.1057 466.6009 .4686 .7769XIMPULDE 35.5890 465.1584 .4257 .7772RESPADAP 36.3010 464.2617 .3718 .7784DECICURI 35.0867 477.7416 .2469 .7856XCAREFRE 35.4152 463.5515 .4247 .7766XJUDGING 52.1652 236.9432 .8948 .7022PERCEIVIa = 0.7848
25.3152 223.1693 .8861 .7167
Note. MBTI continuous variables are italicized.
Table 8Classical Reliability Statistics forJudging-Perceiving Scale (n=420, v=9)
SCALEMEANIF ITEMDELETED
SCALEVARIANCEIF ITEMDELETED
CORRECTEDITEM-TOTAL
CORRELATION
ALPHAIF ITEMDELETED
XFLEXORG 33.4738 61.9062 .4710 .7502PROMPTFR. 33.2200 58.6314 .5701 .7341XRANDSEQ 33.6438 60.9403 .5552 .7380TIMELYRE 33.2224 60.5984 .5251 .7418XIMPETTA 33.9700 65.2885 .4560 .7535XIMPULDE 33.4533 63.7961 .4429 .7544RESPADAP 34.1652 63.7136 .3606 .7682DECICURI 32.9510 67.8656 .2873 .7751XCAREFRE 33.2795 62.9391 .4471 .7538a = 0.7739
Table 9Principal Components Analysis
of 59 PPSDQ Variables and 8 MBTI Continuous Scale Scores(n = 420, v = 67)
FACTOR 1 FACTOR 2 FACTOR 3 FACTOR 4
MIXERLON .09798 .76580 .12198 .17371
SOCPRIVA .16185 .72669 .10241 .16708
INTREXTR -.14694 -.73497 -.04102 -.08167SILENGAB .00604 -.69603 .01098 .08756
PERSNSHY .06954 .76130 .07283 .01396
QUIETEXP -.14331 -.71356 .02916 .15914
GREGARTI .11608 .61047 -.08241 -.14468CONGRECL .04493 .58131 .22576 .04900FRTEDIST .00598 .54446 .33912 .24408SOLIAMIC -.01635 -.57541 -.16895 -.11101EXUBSERE .01839 .52256 -.07160 -.16903STILLANI -.25138 -.46612 -.10626 .22697REFLECAC -.07017 -.39998 .11672 -.30092APPROACH -.14874 .33052 .25744 .20265TERSEWOR .04411 -.20887 -.16069 .23533
EXTRAVER -.08734 -.83588 -.07625 -.05765INTROVER .07975 .82202 .03115 .04266TRADCREA -.39114 -.20472 -.07130 .27441
PRECIMAG -.50480 -.07402 -.26647 .28183
INVENORG .65163 .07157 .06847 -.22385PLANVISI -.52907 -.10889 -.12582 .28745CONCLEXP -.46539 -.09282 -.09253 .1597INSIGHTS .46938 .12253 .19690 -.26435DIVERCON .53006 .22126 .13397 -.10605REALINTU -.24074 .00661 -.26714 .40846DIVERPRE .53842 .21103 .17220 -.17612CONCEPRE .20290 -.17531 -.02025 -.49527DIRECTIN -.32901 -.18801 -.06996 .28574
PRACTHEO -.22529 .08879 .07723 .42507VARIREPI .40447 .25618 .09087 -.01536INQUICRI .30967 .13532 .32427 .00304
SENSING .56729 .07476 .13743 -.52622INTUITIO -.55627 -.01079 -.14620 .50692FACTCOMP -.15102 -.10653 -.68636 .14986TENDERRA .13695 .00378 .61299 -.13407FEELTHIN .01319 .14784 .58127 -.02405KINDANAL .02772 .09428 .60442 .28266
STRICTFO -.25264 -.07131 -.51267 -.18560DISPASEM .07436 -.18164 -.56144 .18529
SKEPTRUS -.13781 -.06767 -.50433 -.32860EMPATHLO .10078 -.09251 .49878 -.24188LOGICHUM -.18645 .01949 -.47804 .14341
LIGHTHEA .31251 .13241 .42186 .07547
GULLSUSP .02779 -.06599 .46025 .11366
CARICOOL -.09240 .19384 .49958 -.05781ACCEPDIS .16384 .17598 .42736 .30983
RECEPTSE .16269 .15654 .38515 .04387
EVALNONJ -.31397 .05331 -.38097 -.14301SYMPATHY -.05152 -.05087 .34039 -.10590TUSTHARM -.17067 .07609 -.33016 -.03317
2 22
EVALOPEN -.24782 -.17206 -.27107 -.16789PRINCIPL -.15305 -.20917 -.29774 -.19792IMPERPER .04478 -.24421 -.37489 .08658SENSUALI .09831 .10012 .23970 -.07497THINKING .12048 -.02762 .83890 -.04635FEELING -.09248 .05752 -.78199 .10323FLEXORGA .63006 .01380 .01942 .06299PROMPTFR -.66313 -.13301 -.05397 -.02819RANDSEQU .62748 .01717 .07887 .02455TIMELYRE -.61195 -.01230 -.09535 -.22273IMPETTAS .51962 -.01700 .06819 -.09822IMPULDEL .48408 .11276 .10313 .05960RESPADAP -.44436 .12569 .07557 .10697DECICURI -.27679 .03029 -.20847 .09560CAREFREE .54022 .04760 .32774 .29978JUDGING .82719 .00185 .01155 .01741PERCEIVI -.81741 -.00522 -.00480 .00511
Note. MBTI continuous variables are italicized.
2326
Table 10Principal Components Analysis
of 59(n =
FACTOR 1
PPSDQ Variables420, v = 59)
FACTOR 2 FACTOR 3 FACTOR 4
MIXERLON -.07032 .75540 .21454 .02571SOCPRIVA -.11309 .71574 .22049 -.01138INTREXTR .13523 -.72812 -.14890 .05664SILENGAB .04450 -.68949 .10593 -.08007PERSNSHY -.07752 .76974 .06533 .09001QUIETEXP .20472 -.70916 .06775 -.01313GREGARTI -.17550 .61145 -.13132 .01986CONGRECL -.02234 .59453 .18430 .17331FRIEDIST .07018 .55307 .36338 .20952SOLIAMIC -.03486 -.58697 -.22764 -.07342EXUBSERE -.06263 .53461 -.19884 .08578STILLANI .32589 -.45540 -.01672 -.09848REFLECAC -.02597 -.41319 -.18822 .29224APPROACH .22028 .34494 .25565 .15171TERSEWOR .04993 -.19459 .11536 -.28967TRADCREA .47390 -.19045 -.03606 -.05335PRECIMAG .58705 -.04192 -.09918 -.26782INVENORG -.69738 .05556 .00547 .09530PLANVISI .61890 -.08268 -.02538 -.12046CONCLEXP .51901 -.07102 -.08584 -.05203INSIGHTS -.54221 .11418 .01032 .27842DIVERCON -.55267 .20764 .13174 .10447REALINTU .37666 .05136 .01926 -.31183DIVERPRE -.58661 .20629 .06512 .19925CONCEPRE -.35126 -.20992 -.27520 .14425DIRECTIN .43930 -.18020 .05075 -.09956PRACTHEO .34682 .12544 .24447 -.04115VARIREPI -.40211 .27768 .14916 .03531INQUICRI -.30391 .13571 .25019 .24203FACTCOMP .20836 -.07409 -.34457 -.59425TENDERRA -.20865 -.02944 .25565 .57298FEELTHIN -.02052 .12887 .25461 .54564KINDANAL .06427 .08187 .55299 .36125STRICTFO .17033 -.06206 -.58296 -.18935DISPASEM .00494 -.14713 -.16164 -.59188SKEPTRUS .00068 -.05327 -.64388 -.14395EMPATHLO -.18588 -.12515 .10099 .55310LOGICHUM .23452 .05007 -.22230 -.42708LIGHTHEA -.28174 .11178 .36008 .27610GULLSUSP .03623 -.09735 .39808 .28084CARICOOL .07377 .20000 .22319 .48937ACCEPDIS -.04556 .17572 .55561 .12681RECEPTSE - .12008 .14157 .34956 .23021EVALNONJ .25:141 .07324 -.52419 -.06660SYMPATHY .02470 -.05004 .01116 .42314JUSTHARM .14046 .10733 -.33184 -.16843'ALOPEN .18093 -.14378 -.40647 -.03346
PRINCIPL .08730 -.20291 -.40177 -.07707IMPERPER -.00059 -.22780 -.14129 -.41645SENSUALI -.13708 .06940 .02618 .34332
24
2i
FLEXORGA -.56157 .01137 .15353 -.04255PROMPTFR .62710 -.12000 -.20690 .09418RANDSEQU -.59109 .00427 .17845 -.02074TIMELYRE .49096 -.01408 -.39844 .19470IMPETTAS -.52884 -.03366 .04728 .04724IMPULDEL -.46839 .11088 .15586 .02414RESPADAP .46962 .14913 -.01687 .14737DECICURI .30133 .04129 -.12709 -.16114CAREFREE -.41680 .03308 .57972 -.01717
Table 11LISREL Maximum-Likelihood Confirmatory Factor Analysis
of 8 MBTI Continuous Scores and 4 PPSDQ Summated Scale Scores
LAMBDA XEXTRINTR
(n=420,
SENSINTU
v=12)
THINFEEL JUDGPERCThetaDelta
FINTREXT 0.782 0.000 0.000 0.000 0.389EXTRAVER -0.990 0.000 0.000 0.000 0.021INTROVER 0.955 0.000 0.000 0.000 0.088FSENSINT 0.000 0.680 0.000 0.000 0.538SENSING 0.000 -0.962 0.000 0.000 0.075INTUITIO 0.000 0.929 0.000 0.000 0.138FTHINFEE 0.000 0.000 0.746 0.000 0.443THINKING 0.000 0.000 -0.982 0.000 0.036FEELING 0.000 0.000 0.894 0.000 0.201FJUDGPER 0.000 0.000 0.000 0.707 0.500JUDGING 0.000 0.000 0.000 -0.984 0.031PERCEIVI 0.000 0.000 0.000 0.980 0.040
PHI
EXTRINTRSENSINTUTHINFEELJUDGPERC
EXTRINTR
1.000-0.086- 0.1060.117
SENSINTU
1.0000.2620.515
THINFEEL
1.0000.172
JUDGPERC
Note. MBTI continuous variables are italicized.
1.000
EXTRAVER
INTROVER
SENSING
INTUITIO
THINKING
FEELING
JUDGING
PERCEIVI
FINTREXT
FSENSINT
FTHINFEE
FJUDGPER
FSCORE1
FSCORE2
FSCORE3
FSCORE4
Table 12
Pearson r Coefficients for 8 MBTI
Continuous Scale Scores
with PPDSQ Summated Scale and Factor Scores
(n = 420)
EXTRAVER INTROVER SENSING
1.0000
INTUITIO THINKING FEELING
JUGGING
PERCEIVI FINTREXT FSENSINT FTHINFEE FJUDGPER FSCORE1
FSCORE2
-.9449**
1.0000
-.0843
.0894
1.0000
.0503
-.0161
-.8955**
1.0000
-.0954
.0770
.2426**
-.2187**
1.0000
-.0480
-.2165**
.2280**
-.8778**
1.0000
.1011*
-.1076*
.0'33
.4653**
-.4711**
.1624**
-.1658**
1.0000
.1262**
-.0872
-.4537**
.4868**
-.1423**
.1847**
-.9647**
1.0000
.1102*
-.7730**
.7475**
.1249*
-.0791
.1085*
-.0751
-.1134*
1.0000
.1646**
-.1602**
-.6553**
.6054**
-.2653**
.2154**
-.5326**
.5292**
-.2530w*
1.0000
.2320**
-.1881**
-.2327**
.2340**
-.7325**
.6584**
-.2052**
.1947**
-.2997**
.3854**
1.0000
-.2463**
.2307**
.1435**
-.1335**
-.4165**
.4250**
-.6951**
.6911**
-.1631**
.6264**
.3664**
1.0000
-.6700**
.6677**
.0828
-.0886
-.6140**
-.1260**
.1114*
-.1004*
.8897**
.2055**
.8307**
1.0000
.0530
.5831**
.0066
-.1508**
-.0130
-.7520**
.7324**
-.0466
.0743
.0149
-.0171
.9637**
-.1218*
.0000
1.0000
.1422**
-.1068*
.0703
-.0155
.0074
.4777**
-.3689**
.1700**
-.1424**
-.0348
-.6767**
-.3503**
.0000
.0000
.5677**
-.5683**
-.6720**
-.0454
.0140
.1476**
-.1614**
-.0763
.0648
.0958*
-.2725**
.0626
.0000
.0000
FSCORE3
FSCORE4
1.0000
.0000
1.0000
Note.
Selected coefficients bearing upon
concurrent validity have been
underlined.
PPSDQ
summated scale scores are named
"FINTREXT" to "FJUDGPER".
3127
BE
ST
CO
PY
AV
AIL
AB
LE
Table 13Spearman rho Coefficients for 8 MBTI Continuous Scale Scores
with PPDSQ Summated Scale and Factor Scores(n = 420)
INTROVER -.9407
SENSING -.0783 .0830
INTUITIO .0477 -.0151 -.9078
THINKING -.0901 .0656 .2505 -.2140
FEELING .0980 -.0404 -.2223 .2208 -.8880
JUDGING -.1009 .0825 .4635 -.4602 .1637 -.1657
PERCEIVI .1268 -.0851 -.4501 .4734 -.1458 .1830 -.9658
FINTREXT -.7720 .7411 .1137 -.0698 .1170 -.0824 .1029 -.1126
FSENSINT .1772 -.1720 -.6555 .6000 -.2659 .2267 -.5235 .5200 -.2768
FTHINFEE .2251 -.1747 -.2293 .2272 -.6908 .6306 -.2214 .2164 -.3075 .4070
FJUOGPER .1331 -.1281 -.4160 .4197 -.2303 .2355 -.6990 .6988 -.1426 .6052 .3738
FSCORE1 .1071 -.1119 -.6178 .5840 -.1492 .1468 -.6810 .6800 -.1221 .8721 .2527 .8297
FSCORE2 -.7545 .7315 .0483 .0035 -.0265 .0581 .0196 -.0289 .9623 -.1725 -.1455 -.0058 -.0208
FSCORE3 -.1167 .0728 .0032 -.0266 .4594 -.3773 .1771 -.1626 .1614 -.1024 -.6712 -.3697 -.0782 .0293
FSCORE4 -.0392 .0015 .1687 -.1682 .5476 -.5570 -.0472 .0435 .1078 -.2880 -.6546 .0210 -.0353 .0153 .0369
EXTRAVER INTROVER SENSING INTUITIO THINKING FEELING JUOGING PERCEIVI FINTREXT FSENSINT FTHINFEE FJUOGPER FSCORE1 FSCORE2 FSCORE3
Note. Selected coefficients bearing upon concurrent validity have been underlined. PPSDQ
summated scale scores are named "FINTREXT" to "FJUDGPER".
3kJ'
28
Figure 1"Scree" Plot for R-technique Factor Analysis
of 59 PPSDQ Variables and 8 MBTI Continuous Scale Scores(n = 420)
11.741 + *I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
E I
G IE IN 6.318 +VAL I
E IS 4.899 +
I
I
I
I
I
I
I
I
I
2.401 +I
2.010 +1.825 + * *
I
1.241 + * * * *
.994 + * * * * * *
.722 + * * * * * * * * * *
.482 + * * *
.270 +
.000 +-+-+-+-+-+-+-4--+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
Figure 2"Scree" Plot for R-technique Factor Analysis
of 59 PPSDQ Variables(n = 420)
9.809 + *I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
E I
1 I
G I
E I
N I
I
A 4.935 + *
L I
U I
E I
S I
3.814 + *
I
I
I
I
I
I
I
I
2.183 + *
1.876 + *
1.637 + * *
I
1.207 + * * * *
1.013 + * * * *
.833 + * * * * *
.615 + * * * * * * * * * *
.425 +
.211 +
.000 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-4.-+-+-4.-+-+-+-+1 2 3 4 5 6 7 8 9 0 1 2 3.4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0
Appendix ADescriptive Statistics
for 8
VARIABLE
MBTI Continuous
MEAN ST. DEV.
Scores and 4 PPSDQ Summated Scale Scores(n=420, v=12)
SKEWNESS KURTOSIS MINIMUM FREQ. MAXIMUM FREQ.FINTREXT 50.284 13.590 0.229 -0.252 16.000 1 88.000 1
EXTRAVER 14.162 6.749 -0.177 -0.934 0.000 1 27.000 5INTROVER 13.183 7.155 0.247 -0.919 0.000 5 29.000 1
FSENSINT 61.840 11.600 0.040 0.032 32.000 2 98.000 1
SENSING 12.633 7.821 0.450 -0.546 0.000 9 34.000 1
INTUITIO 11.969 6.178 0.115 -0.969 0.000 4 25.000 4
FTHINFEE 96.352 15.767 -0.423 0.469 44.000 1 141.000 1
THINKING 10.831 7.742 0.604 -0.466 0.000 16 33.000 1
FEELING 10.245 5.304 -0.084 -0.940 0.000 10 21.000 1
FJUDGPER 37.672 8.796 -0.078 -0.248 12.000 1 61.000 2JUDGING 12.357 7.421 0.145 -0.919 0.000 16 28.000 4
PERCEIVI 14.493 7.940 0.084 -0.921 0.000 11 30.000 4
Note. MBTI continuous variables are italicized.
36
31